KseniiaCheloshkina's picture
Update README.md
5612aa5 verified
metadata
language:
  - en

Artemis-Extactor-70B-4bit-gptq-v1

Artemis-Extactor models aim to extract entities or attributes based on descriptions from document pages, create retrieval queries for search engine or group extracted entities.

Model Details

Model Description

  • Model type: LLM
  • Finetuned from model [optional]: LAMA-3.1-70b

Inference

*.safetensors models can be run using transformers lib or vllm inference engine

Extraction Data

Expected input:

You are an expert in extracting information from document page.

Instructions:
1. If page does not contain entity, you should answer "Not mentioned".
2. If page contains entity, return exact tokens from the pages.
3. Provide json valid structure for the result.
4. Do not add any comment to the answer - it will make json structure valid.

Entity:

Entity Name: Company Name
Entity Description: Name of the company
Entity value: Nokia; Zero Systems

Page Text:
The company is called Apple.

Expected Output:

["Apple"]